A Post-Filtering Approach Based on Locally Linear Embedding Difference Compensation for Speech Enhancement.
Yi-Chiao WuHsin-Te HwangSyu-Siang WangChin-Cheng HsuYu TsaoHsin-Min WangPublished in: INTERSPEECH (2017)
Keyphrases
- locally linear embedding
- speech enhancement
- dimensionality reduction
- noisy environments
- manifold learning
- noise reduction
- low dimensional
- high dimensional data
- high dimensional
- signal to noise ratio
- single channel
- speech signal
- dimensional data
- linear prediction
- principal component analysis
- data points
- vocal tract
- multiresolution
- speech recognition
- wiener filter
- neural network
- computer vision
- geodesic distance
- language model
- nearest neighbor
- pattern recognition
- data structure
- image processing